Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

DM-31116: Adjust backoff criterion for S3 #556

Merged
merged 3 commits into from
Aug 11, 2021
Merged

DM-31116: Adjust backoff criterion for S3 #556

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f6f27d3
Do not include client errors in all retryable list
timj Aug 10, 2021
b7f43dc
Switch size/remove/exists to use io errors not client errors for backoff
timj Aug 10, 2021
6381fd8
For remove intercept NoSuchKey
timj Aug 10, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
19 changes: 14 additions & 5 deletions python/lsst/daf/butler/core/_butlerUri/s3.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,12 +79,18 @@ def on_exception(func: Callable, *args: Any, **kwargs: Any) -> Callable:
RequestError, HTTPError,
# built-ins
TimeoutError, ConnectionError)

# Client error can include NoSuchKey so retry may not be the right
# thing. This may require more consideration if it is to be used.
retryable_client_errors = (
# botocore.exceptions
ClientError,
# built-ins
PermissionError)
all_retryable_errors = retryable_client_errors + retryable_io_errors

# Combine all errors into an easy package. For now client errors
# are not included.
all_retryable_errors = retryable_io_errors
max_retry_time = 60


Expand All @@ -100,7 +106,7 @@ def client(self) -> boto3.client:
# Defer import for circular dependencies
return getS3Client()

@backoff.on_exception(backoff.expo, retryable_client_errors, max_time=max_retry_time)
@backoff.on_exception(backoff.expo, retryable_io_errors, max_time=max_retry_time)
def exists(self) -> bool:
"""Check that the S3 resource exists."""
if self.is_root:
Expand All @@ -109,7 +115,7 @@ def exists(self) -> bool:
exists, _ = s3CheckFileExists(self, client=self.client)
return exists

@backoff.on_exception(backoff.expo, retryable_client_errors, max_time=max_retry_time)
@backoff.on_exception(backoff.expo, retryable_io_errors, max_time=max_retry_time)
def size(self) -> int:
"""Return the size of the resource in bytes."""
if self.dirLike:
Expand All @@ -119,14 +125,17 @@ def size(self) -> int:
raise FileNotFoundError(f"Resource {self} does not exist")
return sz

@backoff.on_exception(backoff.expo, retryable_client_errors, max_time=max_retry_time)
@backoff.on_exception(backoff.expo, retryable_io_errors, max_time=max_retry_time)
def remove(self) -> None:
"""Remove the resource."""
# https://github.com/boto/boto3/issues/507 - there is no
# way of knowing if the file was actually deleted except
# for checking all the keys again, reponse is HTTP 204 OK
# response all the time
self.client.delete_object(Bucket=self.netloc, Key=self.relativeToPathRoot)
try:
self.client.delete_object(Bucket=self.netloc, Key=self.relativeToPathRoot)
except (self.client.exceptions.NoSuchKey, self.client.exceptions.NoSuchBucket) as err:
raise FileNotFoundError("No such resource: {self}") from err

@backoff.on_exception(backoff.expo, all_retryable_errors, max_time=max_retry_time)
def read(self, size: int = -1) -> bytes:
Expand Down